FLASHZIP file Compressor!

[Nania Francesco]

WinFlashzip GUI 1.1.3 is a file archiver and folders, currently with Windows GUI through the huge performance in the compression of all file types. It uses a compression method that I implemented in speed (ROLZ) with arithmetic coding. Retains remarkable speed in compression and decompression. It uses little memory in the minimum configuration with excellent results with all types of files. Enabled since version 1.0.0 the creation of self extraction Archive with which I intended to ensure compatibility for those who want to use it as an archiver. Adding also the possibility in the DOS version to use the extra compression with the switch "-mxn" (where n is a number from 0 to 3) as well as to manipulate the size Dictionary "-kn" (from 0 to 7). Added support for multiprocessing with the ability to use as many as 8 cores. Connecting the extension with the program WFlaship can directly open your files and launch their content.

Only from:
http://heartofcomp.altervista.org/Flashzip.htm

[pat357]

Thx for the nice program !
Is the algo assymmetric in all modes ?

Quick test on enwik9 :

timer flashzip enwik9 enwik9_flashzip09.flashzip

Timer 3.01 Copyright (c) 2002-2003 Igor Pavlov 2003-07-10
Flashzip v.0.9 copyright (C) 2008 By Nania Francesco Antonio
compress... enwik9 1000000000 to 277308041
Kernel Time = 2.854 = 00:00:02.854 = 3%
User Time = 74.459 = 00:01:14.459 = 83%
Process Time = 77.314 = 00:01:17.314 = 86%
Global Time = 89.232 = 00:01:29.232 = 100%

timer flashzip -m1 -s3 -b5 enwik9 enwik9_flashzip09_m1_s3_b5.flashzip

Timer 3.01 Copyright (c) 2002-2003 Igor Pavlov 2003-07-10
Flashzip v.0.9 copyright (C) 2008 By Nania Francesco Antonio
compress... enwik9 1000000000 to 257561029
Kernel Time = 2.714 = 00:00:02.714 = 2%
User Time = 100.308 = 00:01:40.308 = 97%
Process Time = 103.023 = 00:01:43.023 = 99%
Global Time = 103.210 = 00:01:43.210 = 100%

timer flashzip -m2 -s3 -b5 enwik9 enwik9_flashzip09_m2_s3_b5.flashzip

Timer 3.01 Copyright (c) 2002-2003 Igor Pavlov 2003-07-10
Flashzip v.0.9 copyright (C) 2008 By Nania Francesco Antonio
compress... enwik9 1000000000 to 253430870
Kernel Time = 2.870 = 00:00:02.870 = 1%
User Time = 179.541 = 00:02:59.541 = 98%
Process Time = 182.411 = 00:03:02.411 = 99%
Global Time = 182.537 = 00:03:02.537 = 100%

timer flashzip -m2 -s5 -b5 enwik9 enwik9_flashzip09_m2_s5_b5.flashzip

Timer 3.01 Copyright (c) 2002-2003 Igor Pavlov 2003-07-10
Flashzip v.0.9 copyright (C) 2008 By Nania Francesco Antonio
compress... enwik9 1000000000 to 238624750
Kernel Time = 3.978 = 00:00:03.978 = 0%
User Time = 451.170 = 00:07:31.170 = 98%
Process Time = 455.148 = 00:07:35.148 = 99%
Global Time = 456.943 = 00:07:36.943 = 100%

-
Pat357

[Nania Francesco]

-M2 -S7 -B5
name=C:\test\A10.jpg 842468->828574
name=C:\test\AcroRd32.exe 3870784->1482647
name=C:\test\english.dic 4067439->751599
name=C:\test\FlashMX.pdf 4526946->3730272
name=C:\test\FP.LOG 20617071->609216
name=C:\test\MSO97.DLL 3782416->1908321
name=C:\test\ohs.doc 4168192->830446
name=C:\test\rafale.bmp 4149414->1036152
name=C:\test\vcfiu.hlp 4121418->680877
name=C:\test\world95.txt 2988578->574632
number of files=10 original=53134726 compressed=12432736 ENC=38.359001 sec. DEC=2.172000 sec.

[Simon Berger]

Looks like a very strong (ratio) compression program. Gratulations!

[pat357]

@ Nania Francesco
For MOC 2007 it seems you have used QuickLZ v1.40 beta 5 to test and for MOC 2008 you test with version 1.31...
Is it for a specific reason that you went back from 1.40b5 to 1.31 for MOC 2008 ?
Did you maybe have problems with 1.40 bxx ?
Have you ever tested QuickLZ 1.40 b9 ?

I was just wondering why ....

Thanks in advance and keep up the good work !!

[LovePimple]

Thanks Francesco!

[Lasse Reinhold]

Hi pat357

The quick*.exe programs will be replaced by a much more powerfuull archiver qpress.exe.

The reason is that the archiver needs its own version numbering and thus its own name (cannot just follow the core library version numbering anymore).

Current qpress 0.1 is using the QuickLZ 1.40 b9 library

Looking forward to seeing benchmarks, Nania!

(and sorry for being offtopic in a FLASHZIP thread)

[Nania Francesco]

Excuse Me... Qpress and Bit archiver in testing......With top priority

[Nania Francesco]

RAR VS FLASHZIP

[SFC TEST]
flashzip.exe -m1 -s2
number of files=10 original=53134726 compressed=13131165 ENC=3.641000 sec. DEC=2.453000 sec.
rar.exe a -m1
number of files=10 original=53134726 compressed=15399324 ENC=2.858000 sec. DEC=1.545000 sec.
flashzip.exe -m1 -s5
number of files=10 original=53134726 compressed=12818420 ENC=7.045000 sec. DEC=2.234000 sec.
rar.exe a -m2
number of files=10 original=53134726 compressed=13611687 ENC=7.983000 sec. DEC=1.468000 sec.
flashzip.exe -m2 -s1
number of files=10 original=53134726 compressed=13146213 ENC=6.002000 sec. DEC=2.765000 sec.
rar.exe a -m3
number of files=10 original=53134726 compressed=13086728 ENC=10.736000 sec. DEC=1.345000 sec.
flashzip.exe -m2 -s3
number of files=10 original=53134726 compressed=12729384 ENC=9.174000 sec. DEC=2.313000 sec.
rar.exe a -m4
number of files=10 original=53134726 compressed=12364899 ENC=11.016000 sec. DEC=4.640000 sec.
flashzip.exe -m2 -s7
number of files=10 original=53134726 compressed=12446302 ENC=37.611000 sec. DEC=2.188000 sec.
rar.exe a -m5
number of files=10 original=53134726 compressed=12362221 ENC=12.032000 sec. DEC=5.656000 sec.

[CALGARY CORPUS]
flashzip.exe -m1 -s2
number of files=18 original=3251493 compressed=1019948 ENC=0.423000 sec. DEC=0.422000 sec.
rar.exe a -m1
number of files=18 original=3251493 compressed=1170590 ENC=0.656000 sec. DEC=0.439000 sec.
flashzip.exe -m1 -s5
number of files=18 original=3251493 compressed=982764 ENC=0.689000 sec. DEC=0.358000 sec.
rar.exe a -m2
number of files=18 original=3251493 compressed=976651 ENC=0.845000 sec. DEC=0.452000 sec.
flashzip.exe -m2 -s2
number of files=18 original=3251493 compressed=1010902 ENC=0.624000 sec. DEC=0.376000 sec.
rar.exe a -m3
number of files=18 original=3251493 compressed=957232 ENC=0.997000 sec. DEC=0.442000 sec.
flashzip.exe -m2 -s5
number of files=18 original=3251493 compressed=970443 ENC=1.111000 sec. DEC=0.374000 sec.
rar.exe a -m4
number of files=18 original=3251493 compressed=792452 ENC=1.312000 sec. DEC=1.000000 sec.
flashzip.exe -m2 -s7
number of files=18 original=3251493 compressed=970791 ENC=2.141000 sec. DEC=0.343000 sec.
rar.exe a -m5
number of files=18 original=3251493 compressed=790272 ENC=1.454000 sec. DEC=1.139000 sec.

[UCLC HALF LIFE MOD]
flashzip.exe -m2 -s7
number of files=1945 original=129401669 compressed=35222142 ENC=111.531998 sec. DEC=19.813999 sec.
rar.exe a -m3
number of files=1945 original=129401669 compressed=32522973 ENC=75.170998 sec. DEC=43.469002 sec.

[UCLC OTHELLO]
flashzip.exe -m2 -s7
number of files=25 original=5992832 compressed=3460116 ENC=4.718000 sec. DEC=1.032000 sec.
rar.exe a -m3
number of files=25 original=5992832 compressed=3601495 ENC=1.734000 sec. DEC=0.813000 sec.
flashzip.exe -m2 -s2
number of files=25 original=5992832 compressed=3501620 ENC=1.293000 sec. DEC=1.097000 sec.
rar.exe a -m1
number of files=25 original=5992832 compressed=3786275 ENC=1.094000 sec. DEC=0.718000 sec.
[UCLC WAVE AUDIO]
flashzip.exe -m2 -s7
number of files=16 original=51014416 compressed=35727922 ENC=3.625000 sec. DEC=4.843000 sec.
rar.exe a -m5
number of files=16 original=51014416 compressed=35134988 ENC=11.782000 sec. DEC=2.611000 sec.
[UCLC FILE UT2]
flashzip.exe -m2 -s7
number of files=1 original=18203215 compressed=2834510 ENC=15.234000 sec. DEC=0.484000 sec.
rar.exe a -m3
number of files=1 original=18203215 compressed=2899873 ENC=3.140000 sec. DEC=0.313000 sec.

[Black_Fox]

Thanks, has been tested

[Nania Francesco]

Impressive compression for new FLASHZIP 0.91A file compressor!
download link:
http://heartofcomp.altervista.org/FLASHZIP/home.htm

[LovePimple]

Thanks Francesco!


Quick test...

Settings: -m2 -b5

A10.jpg > 833,824
AcroRd32.exe > 1,396,143
english.dic > 790,836
FlashMX.pdf > 3,734,738
FP.LOG > 572,597
MSO97.DLL > 1,850,762
ohs.doc > 801,288
rafale.bmp > 1,142,600
vcfiu.hlp > 633,802
world95.txt > 600,503

Total = 12,357,093 bytes


Settings: -m2 -b5 -s7

A10.jpg > 833,824
AcroRd32.exe > 1,375,093
english.dic > 714,636
FlashMX.pdf > 3,714,709
FP.LOG > 553,813
MSO97.DLL > 1,813,993
ohs.doc > 792,207
rafale.bmp > 996,703
vcfiu.hlp > 623,318
world95.txt > 558,930

Total = 11,977,226 bytes

Compression times were very quick for all settings.

[Nania Francesco]

flashzip.exe -m1 -s1 -b1 C:\calgary\calgary.tar t.zip 3152896->942641
number of files=1 original=3152896 compressed=942641 ENC=0.235000 sec. DEC=0.203000 sec.

rar.exe a -m1 -md4096 t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->1124022
number of files=1 original=3152896 compressed=1124022 ENC=0.219000 sec. DEC=0.109000 sec.

nz.exe a -cf -m800m t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->1307958
number of files=1 original=3152896 compressed=1307958 ENC=0.063000 sec. DEC=0.093000 sec.

flashzip.exe -m1 -s7 -b1 C:\calgary\calgary.tar t.zip 3152896->868632
number of files=1 original=3152896 compressed=868632 ENC=0.953000 sec. DEC=0.218000 sec.

rar.exe a -m2 -md4096 t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->921051
number of files=1 original=3152896 compressed=921051 ENC=0.593000 sec. DEC=0.110000 sec.

nz.exe a -cF -m800m t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->1180742
number of files=1 original=3152896 compressed=1180742 ENC=0.203000 sec. DEC=0.281000 sec.

flashzip.exe -m2 -s1 -b1 C:\calgary\calgary.tar t.zip 3152896->936082
number of files=1 original=3152896 compressed=936082 ENC=0.703000 sec. DEC=0.187000 sec.

rar.exe a -m3 -md4096 t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->901666
number of files=1 original=3152896 compressed=901666 ENC=0.782000 sec. DEC=0.109000 sec.

nz.exe a -cd -m800m t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->995653
number of files=1 original=3152896 compressed=995653 ENC=0.203000 sec. DEC=0.078000 sec.

flashzip.exe -m2 -s7 -b1 C:\calgary\calgary.tar t.zip 3152896->856195
number of files=1 original=3152896 compressed=856195 ENC=2.484000 sec. DEC=0.219000 sec.

rar.exe a -m5 -md4096 t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->762178
number of files=1 original=3152896 compressed=762178 ENC=1.187000 sec. DEC=0.844000 sec.

nz.exe a -cD -m800m t C:\calgary\calgary.tarname=C:\calgary\calgary.tar 3152896->909905
number of files=1 original=3152896 compressed=909905 ENC=0.235000 sec. DEC=0.094000 sec.

flashzip.exe -m1 -s1 -b1 [\OTHELLO\UCLC]
number of files=25 original=5992832 compressed=3420659 ENC=1.202000 sec. DEC=0.813000 sec.

rar.exe a -m1 -md4096 [\OTHELLO\ UCLC]
number of files=25 original=5992832 compressed=3786275 ENC=1.059000 sec. DEC=0.722000 sec.

nz.exe a -cf -m800m [\OTHELLO\UCLC]
number of files=25 original=5992832 compressed=3982074 ENC=0.516000 sec. DEC=0.797000 sec.

flashzip.exe -m2 -s7 -b1 [\OTHELLO\ UCLC]
number of files=25 original=5992832 compressed=3346009 ENC=2.765000 sec. DEC=1.139000 sec.

rar.exe a -m3 -md4096 [\OTHELLO\UCLC]
number of files=25 original=5992832 compressed=3601495 ENC=1.857000 sec. DEC=0.799000 sec.

nz.exe a -cO -m800m [ \OTHELLO\UCLC]
number of files=25 original=5992832 compressed=3176411 ENC=12.235000 sec. DEC=3.906000 sec.

flashzip.exe -m2 -s7 -b5 [\halflife\ UCLC]
number of files=1945 original=129401669 compressed=32525752 ENC=127.444000 sec. DEC=20.559999 sec.

rar.exe a -m3 -md4096 [\halflife\ UCLC]
number of files=1945 original=129401669 compressed=32522973 ENC=76.459999 sec. DEC=44.784000 sec.

Nz.exe [\halflife\ UCLC] test Failed in cf,CF,cd,cD methods ?????

[pat357]

The new Flashzip v0.91a seems to crash on some files : I get a message like "Flashzip has stopped working..Windows is checking a solution for the problem".
I tested it multiple times, even on different systems : every time the same crash !
When crashing, Flashzip has written exactly 16.437.248 bytes to disk (100% repeatable), what probably means it crashes on a well defined chunk of data.

@Nania: can you please try this file in -m2 -s7 -b5 mode ?
(it's a TAR in a RAR archive 62 MB) :

http://rapidshare.com/files/138323573/test4.rar

[Nania Francesco]

Thanks for the signaling that the file send me is in download and I submit it to test ! Hi!

[Nania Francesco]

pat357 THANKS for all!
thanks indeed for the help it was a problem of the ring experimental buffer that I used that it went to copy nonexistent parts !
I put the version on-line compatible 0.91B to 100% with the 0.91A!

[osmanturan]

pat357 THANKS for all!
thanks indeed for the help it was a problem of the ring experimental buffer that I used that it went to copy nonexistent parts !
I put the version on-line compatible 0.91B to 100% with the 0.91A!

At a time, I had same problem in BIT, too

[Nania Francesco]

New FLASHZIP 0.91B is on-line for download!
100% compatible with version 0.91A !
Special Thanks for pat357!

[osmanturan]

Nania, can you tell us some technical details? What's your base method: LZP, ROLZ or LZ77? And also, what is your entropy coder: arithmetic, range or huffman?

[LovePimple]

Thanks Francesco!

You forgot to change the displayed version from v.0.91a to v.0.91b.

[Nania Francesco]

Please LovePimple Re-download the file! The file downloaded is the old version!
to Osmanturan FLASHZIP is ROLZ+ARI+RINGBUFFER
USE [4-256] KEYS IN CONTEST ORDER [2 normal - 3 txt]->&0xffff !

[LovePimple]

Please LovePimple Re-download the file!

Got it!

[pat357]

Nz.exe [\halflife\ UCLC] test Failed in cf,CF,cd,cD methods ?????

What happened ?

[Sami]

>Nz.exe [\halflife\ UCLC] test Failed in cf,CF,cd,cD methods ?????
What happened ?

Indeed, I cannot reproduce any error, and there has never been any bugs in those compressors, although I'm not sure if the bugfixes in 0.04 effect this.

I'm unable to reproduce the fast times of flashzip where it supposedly is near nz-cf. Here is my benchmarks:

uclc-halflife.tar (131mb)     c.size  c.time  d.time (user times)
nz004 -cO                 21.985.959  440.2s   47.6s
arc050a -mx               24.945.210  236.6s    7.1s
ccm130c 4                 25.515.163  155.5s  145.3s
7z -mx9 -m0d=48m          25.580.990  158.9s    4.6s
nz004 -co                 26.450.520  119.0s   21.5s
arc050a -m3               28.369.543   49.6s    7.2s
flashzip091b -m2 -s7 -b5  30.829.055  368.9s   22.8s
rar -m5 -mdg              31.038.631  129.1s    5.0s
flashzip091b -m1 -s7 -b5  31.276.335  144.1s   23.0s
rar -m3 -mdg              31.286.638   78.0s    3.5s
nz004 -cD                 32.045.971   28.2s    3.4s
flashzip091b -m1 -s1 -b5  32.591.569   20.0s   17.9s
arc050a -m2               34.202.182   13.7s    3.8s
nz004 -cd                 36.471.003   11.8s    2.2s
rar -m1 -mdg              36.870.736   16.3s    3.2s
arc050a -m1               41.809.874    4.1s    2.7s
nz004 -cf                 44.475.788    2.8s    2.1s

[Nania Francesco]

HALFLIFE MOD TESTED NON IN TAR MODE ! made a will file for file in any folder!

[Sami]

HALFLIFE MOD TESTED NON IN TAR MODE ! made a will file for file in any folder!

Ok. Can you tell me how nz fails on your test (which file fails)?

Flashzip looks good in your tests. I hope you will improve Flashzip performance with tar and other such concentrated data.

[Nania Francesco]

I don't use the file GUI but the console file nz.exe
the comparison is not equal in these : compression cF
.\media\startupvids.txt
.\materials\VGUI\logos\n.txt [go in crash]

naturally I use an automatic program that compares file of origin and destination comparing the two files!

[Nania Francesco]

The file length of this file is '0' ! I believe is perhaps a problem for many compressor-archiver!

[Sami]

The file length of this file is '0' ! I believe is perhaps a problem for many compressor-archiver!

Thanks! I can confirm that compressing a single zero byte file will display an error for all nz-compressors. I cannot reproduce any crash though.

[Nania Francesco]

I confirm how much you say! in crash doesn't go nz.exe but my program in how much compares 0 byte!!!